Solving an Impossible Riddle with Code

In this video I write code to simulate the 100 prisoner riddle problem with python. This riddle was was covered in Veritasium's amazing video here: • The Riddle That Seems ...
The notebook code can be found here: www.kaggle.com/code/robikscub...
Timeline:
00:00 Impossible Riddle
01:00 The Riddle
02:13 Simulating the Boxes
04:53 Random Strategy
11:18 Loop Strategy
15:03 Graph Network
21:06 Longest Loop
Follow me on twitch for live coding streams: / medallionstallion_
My other videos:
Speed Up Your Pandas Code: • Make Your Pandas Code ...
Speed up Pandas Code: • Make Your Pandas Code ...
Intro to Pandas video: • A Gentle Introduction ...
Exploratory Data Analysis Video: • Exploratory Data Analy...
Working with Audio data in Python: • Audio Data Processing ...
Efficient Pandas Dataframes: • Speed Up Your Pandas D...
* KZbin: youtube.com/@robmulla?sub_con...
* Discord: / discord
* Twitch: / medallionstallion_
* Twitter: / rob_mulla
* Kaggle: www.kaggle.com/robikscube
#python #numpy #simulation #riddles

[@ritvikrastogi4912]

Hi Rob, Greetings from India I really love your content and for this tutorial, I can not even put into words how much I liked it I'm really inspired and will try out to simulate more such mathematical problems Thanks a lot for inspiring a lot of individuals like myself

[@robmulla]

Hey Ritvik - I'm so glad you found this tutorial helpful. I think that doing simulations like this can be really fun. I think it's great that you are going to do similar simulations. Please share the link to this video with anyone else you think might enjoy it!

[@H99x2]

After watching that Veritasium video when it was just released I thought to myself "I wish I could code this up in Python.." But since I just started learning it, I couldn't. This video is exactly what I need! Thanks Rob

[@robmulla]

That's awesome! Glad to hear I wasn't the only one thinking this. Please share this with anyone else you think would like to learn how to code it in python.

[@marc-andremaillet5893]

I’m from this video too! I was looking for a computer simulation to solve this riddle for good.

[@robmulla]

@@marc-andremaillet5893 awesome. Hopefully you found the video helpful.

[@marc-andremaillet5893]

@@robmulla Yes thanks!

[@ChaiTimeDataScience]

Im absolutely loving the diversity of tutorials! Thanks Rob!

[@robmulla]

Thanks so much! Trying something somewhat different in this tutorial. I’ve been enjoying your new content too!

[@ThePaintingpeter]

Like Ritvik Rastogi said, these type of problem solving videos are fantastic. They're engaging, challenging and rewarding.

[@robmulla]

Glad you like them! This was probably the most fun I've had making a video because I think the riddle is so interesting.

[@davoodastaraky7608]

I learn a lot from your videos. Thanks for spending time to create these content.

[@robmulla]

This makes me so happy 🤗 thanks for watching.

[@ronbzalen]

Absolutely beautiful! Got to love a good mix of data science and graph theory :)

[@robmulla]

Glad you enjoyed it!

[@md.sakiburrahman4409]

THIS IS LITERALLY THE BEST THING , I ABSOLUTELY LOVE IT, THANK YOU SO MUCH. PLEASE CONTINUE MAKING MORE VIDEOS. YOU HAVE MY SUBSCRIBTION ❤❤

[@robmulla]

I'm so glad you like it. It's personally the favorite video I've made even though it's not the most popular. Thanks!

[@reubenthomas1033]

Amazing! Awesome video, you’ve inspired me to simulate a similar problem myself

[@robmulla]

So happy to hear that. My goal is to inspire viewers to explore what’s possible with data science so I’m especially glad to hear you are going to take action.

[@EVL624]

Hey, when you create the try_strategy() function you loop over the prisoners, but don't actually use the resulting values from the loop. The looping variable is called prisoner, but you are inserting prisoner_number into the strategy() function. The reason you don't get an error is that prisoner_number is still stored in the global scope from when you assigned prisoner_number = 1 earlier.

[@robmulla]

Great catch. Do you think it messed up the output? Can you point me to the timestamp?

[@EVL624]

It should intuitively increase the winning rate slightly since we are repeatedly looking at the same prisoner 100 times, meaning that if prisoner 1 never finds their number the simulation results in a loss (which it should have if it was implemented correctly too), but if prisoner 1 does find their number, the simulation always results in a win since the same condition is just checked 100 times (this would not be the case if it was implemented correctly, since prisoner 1 could find their number, while some other prisoner does not, resulting in a loss). The timestamp is around 7:30 Edit: My reasoning is wrong for the random strategy since you always choose a random box the result should not be impacted. I will have another think about the loop strategy as well.

[@realdriss0]

this video is a gem

[@robmulla]

Thanks 🙌

[@gustavojuantorena]

Great and unexpected video! 👏👏

[@robmulla]

It’s a little different than my other videos. Glad you liked it!

[@louisvuittondonvg9040]

dude you're unbelievable this is so good

[@robmulla]

Thanks 🤗

[@ronin6158]

this really is wild; good to see in hard sim, not just equations.

[@robmulla]

Glad you enjoyed it. Had a lot of fun making this video.

[@Levy957]

Loved the vídeo!!!

[@robmulla]

Glad you liked it Levy! Thanks the comment

[@dangalimov7435]

Deserves to be seen in tops

[@robmulla]

Thanks! I’m happy with this video.

[@TheThunder005]

Great video!

[@robmulla]

Thanks so much for watching. Share with a friend 😀

[@noahcollin9532]

Great content! 👍 How do you get your notebook to tell you about function calls? @5:17 You typed "np.random.choice" and a little window popped up telling you what parameters are expected. How do you enable that functionality? Thanks

[@robmulla]

Thanks for watching! This is built into jupyter you just need to be within the function and type shift-tab. I believe it’s using something called Jedi in the backend you can search about.

[@philtoa334]

Great.

[@robmulla]

Glad you liked this video. It’s one of my favorites that I’ve made even though it’s not very popular.

[@jonasherseth5306]

Very good tutorial. In the for-loop @7:38, shouldn't you pass prisoner to the strategy function instead of prisoner_number?

[@robmulla]

Thanks for the feedback. So glad you liked it. Oh no, did I mess something up? I looked at the time stamp you linked but I think you might be referring to a different part?

[@jonasherseth5306]

@@robmulla in the for loop in the try_strategy function, you have "correct = strategy(prisoner_number, boxes)". I think the should be "correct = strategy(prisoner, boxes"... But I might be mistaken

[@jonasherseth5306]

@@robmulla actually, I see you have corrected this on kaggle to "correct = strategy(prisoner_number=prisoner, boxes=boxes)"

[@robmulla]

@@jonasherseth5306 nice catch. I remember having to edit out a part where I fixed a bug while recording the video so you may have caught it before I did! Thanks for watching

[@ulf594]

Hi in which python environment did you do this? :)

[@robmulla]

Hey. I’m coding in this video in a jupyterlab environment, however I’ve also posted a link to the Kaggle notebook version in the description. Check out my video on jupyter to get more details on my setup.

[@Jakub1989YTb]

6:25 .. just use `return prisoner_number in prisoner_choices` .. You seem to know what you are doing, but it lacks the final pythonic touch. This video is full of small antipatterns.

[@robmulla]

Great points. I’d love any suggestions you have for ways the code could be improved. Looking to learn!

[@Jakub1989YTb]

Why don't you just use `list(range(100))`?

[@robmulla]

Yea that would’ve been cleaner.

[@qwang3118]

The problem is interesting, but the presentation is misleading and confusing. It is NOT that prisoners have 31.2% chance to be freed. 31.2% chance is for the map f: Box Labels - -> Prisoner's # to be good, in the sense that all cycle lengths of f

[@robmulla]

Thanks for your thoughtful and detailed comment. I'm going to need to read it a few more times to fully understand. It sounds like you have a very strong understanding of this problem. I learned it from the channel @veritasium who had a video about it. Have you seen that video? Does it also make the same mistake or is it just something I didn't explain well in my video. Interested to hear your thoughts.

[@snippletrap]

return prisoner_number in prisoner_choices

[@robmulla]

Love it. Much better than my code and more pythonic. It might be less readable for newbies but maybe not.